Journal of Petroleum Science and Engineering 78 (2011) 637–645
Contents lists available at SciVerse ScienceDirect
Journal of Petroleum Science and Engineering journal homepage: www.elsevier.com/locate/petrol
Heavy oil and bitumen recovery by hot solvent injection V. Pathak a, Tayfun Babadagli a,⁎, N.R. Edmunds b a b
University of Alberta, Canada Laricina Energy, Canada
a r t i c l e
i n f o
Article history: Received 15 March 2011 Accepted 8 August 2011 Available online 19 August 2011 Keywords: Hot solvent Heavy-oil/bitumen recovery Propane Butane Saturation temperature Asphaltene precipitation
a b s t r a c t Thermal and miscible methods are commonly used for in situ recovery of heavy oil and bitumen. Both techniques have their own limitations and beneﬁts. However, these methods can be combined by co-injecting solvent with steam or injecting solvent into a pre-heated reservoir. The current work was undertaken to study the performance of solvents at higher temperatures for heavy oil/bitumen recovery. Glass bead packs and Berea sandstone cores were used in the experiments to represent different types of pore structures, porosity and permeability. After saturating with heavy oil, the samples were exposed to the vapor of parafﬁnic solvents (propane and butane) at a temperature above the boiling point of the solvent, and a constant pressure of 1500 kPa. A mechanical convection oven was used to maintain constant temperature across the setup. The setup was designed in such a way that a reasonably long sample (up to 30 cm) can be tested to analyze the gravity effect. The oil recovered from each of these experiments was collected using a speciﬁcally designed collection system and analyzed for composition, viscosity and asphaltene content. The ﬁnal amount of oil recovered in each case (recovery factor but not extraction rate) was also analyzed and the quantity and nature of asphaltene precipitated with each of the tested solvents under the prevailing temperature and pressure of the experiment was reported. Optimal conditions for each solvent type were identiﬁed for the highest ultimate recovery. It was observed that recovery decreased with increasing temperature and pressure of the system for both solvents, and that the best results were found when experimental temperature is only slightly higher than the saturation temperature of the solvent used. It was also noticed that butane diluted the oil more than propane which resulted in lower asphaltene content and viscosity of oil produced with butane as a solvent. © 2011 Elsevier B.V. All rights reserved.
1. Introduction SAGD and VAPEX are the most frequently used techniques for the recovery of heavy oil and bitumen. In both of these techniques, recovery agents, i.e., steam and solvent vapors for SAGD and VAPEX, respectively, are introduced into the reservoir through a horizontal injection well, and displaced oil is produced from a horizontal production well located below the injector. In SAGD, recovery is enhanced due to the heat transfer from steam to the heavy oil whereas in the VAPEX process (originally proposed by Butler and Mokrys, 1991), solvent vapors dissolve at the interface between solvent and heavy oil and diffuses through the oil. In both these processes, the less viscous produced oil ﬂows down by gravity to the producer well. SAGD is not economical in cases where reservoirs are thin, because heat losses to conﬁning strata become excessive compared to the resource. Another problem in SAGD is the cost involved for treating efﬂuent water and the high energy requirements in order to have ⁎ Corresponding author at: Department of Civil and Environmental Engineering, School of Mining and Petroleum Eng., 3–112 Markin CNRL-NREF, Edmonton, AB, Canada T6G 2W2. E-mail address: [email protected]
(T. Babadagli). 0920-4105/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.petrol.2011.08.002
a continuous production of steam. Energy losses incurred in SAGD operations are much higher than in VAPEX (Upreti et al., 2007). Singhal et al. (1996) presented guidelines for making a decision between SAGD and VAPEX for heavy oil reservoirs. They stated that the VAPEX process is highly energy efﬁcient as compared to SAGD. Moreover, condensation of fresh water from steam poses problems to the formation by clay swelling and reduction in oil relative permeability. VAPEX does not pose such problems. SAGD operations also result in large CO2 emissions which can be reduced by using solvent. Lastly, solvents used in VAPEX are either insoluble or sparingly soluble in water, hence, there is no concern of solvent loss to water (Das and Butler, 1996). In the case of SAGD, however, steam may be lost to water. Thus, the use of hydrocarbon solvents seems more promising than using steam alone. Using gaseous solvent is better than liquid ones because it reduces the amount of solvent needed, results in higher rate of diffusion, and provides a high density contrast for the gravity drainage (Friedrich, 2005). The main problem in VAPEX is low initial recovery because of the slow nature of diffusion process. Another problem appears to be high asphaltene precipitation resulting in reduced permeability. This also implies lower recovery of bitumen, although of slightly upgraded quality. Thus, selecting the type of solvent and application conditions becomes a critical issue.
V. Pathak et al. / Journal of Petroleum Science and Engineering 78 (2011) 637–645
When gases like propane and butane are used for VAPEX process, they may condense because of the high pressure in reservoir. Thus, heating of reservoir is needed prior to solvent injection. The hotsolvent technique combines the heating effect of steam or hot water with the dilution effect of solvents to give even a better heavy-oil/bitumen recovery than SAGD or VAPEX alone. This also ensures that the solvent is maintained in vapor phase which enhances gravity drainage and mass transfer, and reduces the solvent inventory held up in the reservoir. This is not a new concept and it was patented more than 25 years ago by Allen et al. (1984). They used superheated pentane injection and showed its effectiveness in tar sand recovery. The core used by them was drilled from a center and a heating rod was placed in this inner bore. Palmgren and Edmunds (1995) tested the effects of injecting high temperature naptha instead of steam and found that recovery was improved by using naptha. Several other researchers studied the performance of the steam injection process if some solvent is added in small proportions (Redford and McKay, 1980; Shu and Hartman, 1988, Goite et al., 2001). A variation of this technique was studied by various researchers and was named Expanding Solvent-SAGD (ES-SAGD). In this technique, a solvent is injected in low quantities along with steam and solvent was chosen such that it would evaporate and condense at the same conditions as water under the prevailing reservoir pressure and temperature (Nasr et al., 2003). A low pressure ES-SAGD process was also tested experimentally and numerically in which the pressure of the system was maintained at 1500 kPa (Ivory et al., 2008). Frauenfeld et al. (2007) conducted experiments on sandpack models with various solvents and different well conﬁgurations and different ways of heating the system. They found that the amount of solvent needed in a thermal VAPEX process was much less than needed in a normal VAPEX process for the same amount of oil recovery. Superheated pentane was used in a warm VAPEX experiment by Rezaei and Chatzis (2007). They found the results to be promising in terms of oil recovery. Currently, several commercial ﬁeld scale projects are underway to investigate the performance of SAGD in the presence of solvents (Edmunds et al., 2009a, 2009b). This technique has a great potential in terms of combining the heat and mass transfer effect to yield a more efﬁcient process as compared to SAGD or VAPEX alone. 2. Statement of problem The hot solvent process is a variant of the VAPEX process with the difference being the incorporation of temperature. Hence, it is a complex process involving both heat transfer (transfer of sensible heat to the oil) and mass transfer (of lighter components in the oil phase). As an example, when light solvents are added to steam in SAGD, the more-volatile solvent forms a vapor zone between the steam front and the solvent condensation zone (Fig. 1). Oil is substantially drained by the time (if ever) the steam reaches a given point. Before
Fig. 1. Schematic representation of heating and condensation zone during hot solvent injection.
the solvent front reaches a given point, the rock will have been prewarmed by thermal conduction. A characteristic of this process is that the amount of solvent that condenses within a given pore volume can be controlled to a fraction (e.g. 10–30 vol.%) of the OIP, by simply controlling the solvent 'dose' in line with the rate of drainage from steam alone. The objective of the current experimental work was to simulate the history of an element of rock that is exposed to such a front, without having to construct and operate a full-up 3D physical model. Key to this is controlling the amount of solvent that can condense on or in the core, by thermodynamic means. The main objective of the current work was to assess the performance of heated hydrocarbon solvent vapors for recovery of heavy oil under a ﬁxed pressure and quantify the expected amount of asphaltene precipitation for each experiment. The solvents used were propane and butane. The heavy oil used was from the Lloydminster area and the bitumen used was from an Albertan ﬁeld. The experiments were aimed at analyzing the recovery factor (or ultimate recovery) for cases where a pure solvent is in contact with heavy oil/bitumen in a heated rock. In other words, all experiments were conducted with pure solvents (selection of propane and butane) and no steam, with the heating of the samples done by a mechanical convection oven. The samples were to be kept soaked in the solvent and not in a regular injection-production scheme. The main purpose was to observe the sensitivity of the process to pressure and temperature, which critically inﬂuence the phase behavior of the solvent and, consequently, asphaltene precipitation and ultimate recovery. 3. Experimental details The details of the experiments for this work were as follows: 3.1. Setup The special considerations for this study were: 1. As a constant temperature was required for the experiment (higher than the boiling temperature of the solvent), two heating options were tried. The ﬁrst check was done with a hot water bath with core holder sitting in it. The thermocouple readings from the hot water bath showed ﬂuctuations and an oven was selected as the medium for experiments. A mechanical convection oven was found to keep the entire system at a constant temperature. 2. The sample was to remain exposed to solvent vapors for a sufﬁciently long time for diffusion and then the products were to be drained into a sample collection system at the end. As the pressure was to be maintained by injecting the gas at a constant pressure, a pressure vessel was designed. The vessel was capable of handling samples up to 30 cm in height to have a meaningful gravity effect. The gas was to be drained only from the production end along with the diluted oil at the end. Therefore, the gas was also at the same temperature as the oven. The experimental setup is shown in Fig. 2. The set-up consisted of a cylinder to house the saturated core or the glass beads sample. The largest sample which the cylinder could ﬁt was 5 cm in diameter and 30 cm in length. This gave us a chance to explore the effect of gravity by changing the height of the sample. The cylinder was placed inside an oven and a thermocouple and a pressure transducer were installed at the center of the cylinder. Hydrocarbon gases were injected using high pressure cylinders and the produced oil was collected in a sealed sample collection system. The hydrocarbon gases were not recovered and vented out through fume hood. Experiments were conducted on glass beads (500 micron diameter) samples mixed with heavy oil obtained from the Lloydminster area. Porosities, i.e. the original oil in-place, were typically in the
V. Pathak et al. / Journal of Petroleum Science and Engineering 78 (2011) 637–645
Fig. 2. Schematic representation of the experimental set-up.
range of 30–40%. A few experiments were done using Berea sandstone cores. The glass beads were selected to represent the porous media as it is easier to control the capillary diameter with uniform sized glass. It also allowed us to understand the effect of gravity alone without worrying about capillary pressure changes due to changes in the capillary diameter. A wire mesh basket was designed for holding the glass bead–heavy oil mixture. The mesh selected was ﬁner than the glass beads. The length of the basket was 30 cm and this gave us the option of testing various lengths of the sample. 3.2. Procedure In all the experiments, a metal funnel was placed in the cylinder to direct the ﬂow of the diluted oil towards the exit. Once the setup was tested for leakages, the oven was started and the system was left to equilibrate overnight. This was done to ensure the steady state condition with a constant temperature. The temperature and pressure were constantly monitored using a thermocouple and a pressure transducer mounted at the center of the cylinder. The data was logged using an acquisition system. After a constant temperature was ensured, the solvent gas was injected in a quantity required to have a pressure of 1500 kPa inside the cylinder. The exit valve was kept closed so that the gas would heat up and the sample was left exposed to the gas for a long time (4 to 48 h, depending on sample size and type), which is termed as soaking time. For glass bead samples, it was decided to have a soaking time from 4–12 h depending upon the volume of oil being used in the experiment. Later, the initial experiments showed that this estimate of soaking time for glass bead experiments is quite reasonable. After that, the exit valve was opened and the diluted oil was collected in the sample collection system. The gas was vented to a fume hood and the oil was taken for analysis. Care was taken to drain the produced oil completely from all the piping. The produced oil was analyzed for asphaltene content, refractive index, viscosity, and, in some cases, for composition.
Two experiments were also conducted with an Alberta bitumen sample. Its properties are listed in Table 2. The bitumen was very viscous and was produced during a ﬁeld scale cold solvent injection process (Edmunds et al., 2009a) and had no components lighter than C9. 4. Results Results for all the experiments conducted with heavy oil are presented in Table 3 and results for the two experiments conducted with bitumen are presented in Table 4. In general, the recovery was observed to decrease with increasing pressure keeping all other parameters constant as can be seen from experiments 3, 4 and 5 in Table 3. Experiments 9 and 10 also showed the same behavior. Temperature was also observed to adversely affect the recovery. Asphaltene content of produced oil is directly related to the weight percentage of original oil left over as asphaltene precipitates in the porous media. In general, asphaltene content and viscosity of oil produced with butane as a solvent were observed to be less than those of the oil produced with propane as a solvent. Recovery was also higher when butane was used as solvent as can be seen from experiments 1 and 2 in Table 4 for bitumen. Detailed analysis of the results is provided in the next section. 5. Analysis of results 5.1. Recovery Various glass-bead experiments showed recovery varying from 48% to 95% (Table 3). Only the overall recovery factors were recorded. The experiments did not involve continuous injectionproduction so there was no continuous recording of production rates. The recovery was found to be highly dependent on the temperature. The recovery was less when the temperature was much higher than the saturation temperature. This is in accordance with the Raoult's law of partial pressures. Raoult's law states that if the
3.3. Heavy oil and bitumen properties The oil used in most of the experiments was a heavy oil from the Lloydminster area in Alberta with no light components (C6 and below). Its properties are given in Table 1. Fig. 3 illustrates the phase plot for this oil showing that it is dead oil, which is also obvious from the composition. The oil had no components lighter than C7, and the ﬁngerprint plot in Fig. 4 implies that it contains mostly the components of C30 or more. The viscosity of the oil was around 9,000 cP at 25 °C and the speciﬁc gravity was 0.96. The asphaltene content, as measured by excess solvent-ﬁlter paper method, was about 14.6% by weight.
Table 1 Properties of tested heavy oil (obtained from the Lloydminster area). Property
Density (at 25 °C) Degree API Viscosity (at 25 °C) C6+ molecular weight Refractive index Asphaltene (% by weight)
0.96 g/cc 15 9231 cP 395 g/mol 1.550 14.6
V. Pathak et al. / Journal of Petroleum Science and Engineering 78 (2011) 637–645 Table 2 Properties of tested Alberta bitumen.
Fig. 3. Phase plot of the dead oil used.
components of a solution are in equilibrium, then the total vapor pressure (ptotal) is given by:
ptotal ¼ xA pA þ xB pB þ :::::
where xA and xB are the mole fractions of components ‘A’ and ‘B’ in liquid, and p⁎A and p⁎B are the vapor pressures of pure component A (propane in our case) and pure component B (tested heavy oil). The tested heavy oil had no light components, hence its vapor pressure can be assumed negligible. This means that,
ptotal ¼ xA pA
As (ptotal) is constant, the quantity of solvent gas in liquid is inversely proportional to its saturation pressure at that temperature. Saturation pressure increases with temperature, making the value of xA smaller. Thus, the solution will have very little quantity of solvent and the dilution effect of the solvent will be minimal. As a result of this, recovery will be less. Therefore, after testing higher temperatures, it was decided to test temperatures in the range of the saturation temperature of the solvent gases. However, it was also seen that recovery was lower for the cases when the experimental temperature was lower than the saturation temperature. The best results were seen when the experimental temperature and pressure were very close to the solvent phase envelope, in the vapor region, as shown in Fig. 5 through Fig. 8. A sample calculation has been shown
Density (at 25 °C) Degree API Viscosity (at 60 °C) Refractive index Asphaltene (% by weight)
1.033 g/cc 5–9 22030 cP 1.592 15
in Appendix A. It should also be noted that the experiments where the solvent is in liquid phase (all experiments to the left of the phase envelope) may give a good recovery, but at the expense of using a larger quantity of solvent and a slower diffusion process. The analysis here considers only those experiments in which solvents were present in vapor phase. The recovery was also observed to be affected by pressure. For example, with an increase in pressure in experiments, 3, 4, and 5 at the same temperature, the recovery was reduced. This can be attributed to the condensation of solvent in the system, which results in slower diffusion of the solvent into the heavy oil (note that the saturation pressure at 98 °C is about 1,500 kPa for butane). Contrary to our expectations, height of the sample did not seem to have a signiﬁcant effect on recovery. It was found that the recovery decreased with the increase in the product of height of the sample, pressure of the system and the difference of the experimental temperature and saturation temperature. Fig. 9 shows an increasing trend of recovery with a decrease in the product of these three physical quantities. Fig. 11 presents the recoveries for the different experiments given in Table 3 showing that, for similar conditions, butane is likely to give better results. A curve ﬁt was also attempted for recovery, assuming that the recovery is a function of sample height, operating temperature, operating pressure, and other factors which need to be explored further. Many functions were tested and the following two were observed to yield the best ﬁt: Recovery ¼ 0:244h−11:903ln ðT−Tsat Þ−0:086P þ 217:619
and, Recovery ¼ 2:885ðh=t Þ−13:727ln ðT−Tsat Þ−0:0999P þ 237:502
Here, h T Tsat P t
Fig. 4. Fingerprint plot for the dead oil used.
Sample height in cm Experimental temperature in °C Saturation temperature for the selected solvent at experimental pressure in °C Experimental pressure Soaking time in hours
The coefﬁcients in Eq. 3 indicate that temperature and pressure have an adverse effect on recovery, while sample height effect is in positive way. This is in accordance with above discussion. The last coefﬁcient shows that there may be other factors dominant in the process too. The ﬁt to the experimental data was shown in Fig. 10. The experiments chosen to achieve the ﬁt are only the ones where the solvent is present in vapor phase. Looking at Fig. 10 and matching experiments in Table 3, experiments 6, 10, 11, 13 and 14 appear to be away from the general trend. These are also the experiments in which the produced oil contained a higher percentage of asphaltenes. In other words, asphaltene precipitation in the porous media was lower than in the other experiments. This infers that the role of asphaltene precipitation on recovery is critical and needs more attention as it directly impacts the drainage rate and phase entrapments (i.e., recovery).
V. Pathak et al. / Journal of Petroleum Science and Engineering 78 (2011) 637–645
Table 3 Summary of experiments with heavy oil. Porous media details S. no.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 17 18
Butane Butane Butane Butane Butane Butane Butane Propane Propane Propane Propane Propane Propane Propane Propane Propane Butane
Glass beads Glass beads Glass beads Glass beads Glass beads Glass beads Glass beads Glass beads Glass beads Glass beads Glass beads Glass beads Glass beads Glass beads Glass beads Berea core Berea core
500 μ 500 μ 500 μ 500 μ 500 μ 500 μ 500 μ 500 μ 500 μ 500 μ 500 μ 500 μ 500 μ 500 μ 500 μ
Asphaltene content (wt.%)
Approx soaking time (h)
29 29 18 26 10 17 17 29 15 17 17 17 23 27 20 15 15
5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5
40 40 30 30 30 30 30 40 40 40 30 30 30 30 30 23 21
70 80 98 98 98 112 108 90 85 67 52 54 53 53 52 53 98
1030 1030 1400 1500 1600 1500 1600 1500 1500 1500 1500 1830 1500 1500 1500 1500 1350
55.6 52.6 94.5 72.1 62.3 45 64.5 55.3 53.7 47.8 83.8 64.2 75.5 60.3 65.5 27.5 44.4
5.7 6.5 6.7 11.3 Not measured 11.3 13.8 13.7 11.4 12.5 10.1 10.6 12.3 13.6 Not measured Not measured Not measured
1.54 1.543 1.536 1.551 1.545 1.555 1.554 1.555 1.554 1.554 1.545 1.545 1.546 1.548 1.544 1.553 1.545
4 4 6 12 8 7 8 4 4 4 4 4 10 10 6 48 28
The constants in Eqs. (3) and (4) imply that there could be other parameters affecting the recovery. A further investigation on other dominant mechanisms was undertaken in another research (Pathak et al., 2011). The Berea core experiments were much more difﬁcult to drain oil and gave much less recovery even if the solvent soaking period was almost 10 times than of the glass bead experiments. Obviously, this was due to much lower permeability of the sandstone than the glass bead models yielding a higher capillarity. Due to the constraint of time, it was decided not to soak the core for more than 2 days (compared to 4–8 h of soaking time for glass bead experiments). For such a soaking time, the recovery was found to be about 28%. To overcome the increased capillarity by decreasing pore size, much longer sandstone samples are needed to drain the oil. 5.2. Asphaltene content of produced oil and asphaltene precipitation/ deposition Asphaltenes are deﬁned as the n-heptane insoluble, toluene soluble components of heavy oil, bitumen or coal (Sirota, 2005). These are the components which increase the viscosity of oil and are also undesirable because they deposit inside tubulars. During conventional VAPEX processes, if parafﬁns are used as solvent, after a certain solvent concentration, asphaltenes start to drop out as precipitates (keeping pressure and temperature constant). Asphaltene precipitation is a very important issue in heavy oil recovery by miscible methods and is still not fully explored. A general perception is that asphaltene precipitation is favorable for recovery as it reduces the oil viscosity and “upgrades” the oil. However, as the asphaltenes get deposited on the pore throats, they adversely affect the permeability. Haghighat and Maini (2008) conducted experiments to investigate the effect of asphaltene precipitation on overall recovery. They found that when operating pressure was less than the vapor pressure of solvent, no asphaltene precipitation occurred because oil was not upgraded. For pressures higher than the vapor
pressure of the solvent, the beneﬁts and disadvantages of asphaltene precipitation almost negated each other. Johnson et al. (1975) tested the effectiveness of 25 solvents based on heavy oil viscosity reduction and preventing asphaltene precipitation. They indicated that solvents with higher aromatic content are likely to give low asphaltene precipitation. Beyond all these, asphaltene is a commercial product and is part of the oil. The ultimate target is to produce it rather than deposited in the reservoir. To measure the asphaltene content of produced oil and precipitation, a small quantity of the produced oil sample (4–5 g) was mixed with a large quantity of n-heptane (40–50 times by weight), and left to stand for 2–3 days. Next, the mixture was ﬁltered using an 11 μ ﬁlter paper, and a vacuum pump was used to assist in increasing ﬁltration rate. The system was covered with aluminum foil to minimize the evaporation losses during ﬁltration. The schematic for the setup is shown in Fig. 11. Then, the ﬁlter paper was removed carefully with all ﬁltered asphaltenes contained in it, and placed in an oven overnight to evaporate all the heptane. Weight of the dried ﬁlter paper with asphaltene gave the asphaltene content of the produced oil. The asphaltene contents of produced oils in various propane and butane experiments are shown in Fig. 12. It shows that experiments with propane yielded diluted oil with asphaltene content higher than the oil obtained from butane experiments. This directly relates to the amount of asphaltene that was deposited from the original oil on the pores (Fig. 13). The asphaltene contents of the produced oil are given in Table 3. A Dean Stark analysis was also done on the sample from experiment 9 and the value of asphaltene precipitation (in the sample after the experiment) was found to be about 3% by weight. This is a bit lower as compared to the value measured using ﬁlter paperexcess solvent method (about 6%). The extraction for Dean Stark process was done in 2 stages, ﬁrst with n-heptane and then with toluene. While the Dean Stark analysis is more time consuming (takes upto 10 days to have a reliable result), the ﬁlter paper-excess solvent method is a more practical approach to measure asphaltene content.
Table 4 Summary of experiments with bitumen. Sample details
Asphaltene content (wt.%)
Approx soaking time (h)
Glass beads 500 μ Glass beads 500 μ
8.2 Not measured
1.555 Not measured
V. Pathak et al. / Journal of Petroleum Science and Engineering 78 (2011) 637–645
Fig. 5. Phase envelope for butane with experimental pressure and temperature indicated by the red points. The numbers indicate the recovery for the corresponding experiment and the numbers in parentheses indicate the experiment number, showing that recovery is higher if the experimental conditions are very slightly on the right of the phase envelope.
Fig. 7. Phase envelope for butane with experimental pressure and temperature indicated by the red points. The numbers indicate the ratio of recovery for the corresponding experiment and sample height (in cm) giving a value of 4–6 for the best experiments. The numbers in parentheses indicate the experiment number.
More Dean Stark analyses will be done at a later stage of the research and the results will be compared to the ﬁlter paper-excess solvent method.
sample introduced into the instrument is very small (and is dissolved in toluene prior to injection) and thereby introduces a problem of detection of trace quantities of the light hydrocarbons. A headspace-GC technique involves an additional instrument which allows a much larger sample size to be heated. The hot headspace gases above the sample can be collected and then injected into the GC. This kind of analysis will be done at later stages of the research to conﬁrm the ﬁndings of GC–MS.
5.3. Composition It was expected that the solvent would diffuse into the heavy oil in experimental conditions and would “upgrade” the oil. This essentially involves a degree of change in composition and viscosity. However, all analyses were made after ﬁnishing the experiments and letting the system equilibrate to standard conditions. Under the standard conditions, all the solvent was expected to leave the liquid phase as the boiling temperatures of both propane and butane are very low under atmospheric pressure. To conform this, a Gas Chromatography–Mass Spectroscopy (GC–MS) measurement of oil produced was performed on two samples, (for propane and butane based experiments). No light fractions were seen in the oil after conducting the GC–MS analysis. Thus, the produced oil is not likely to contain any light hydrocarbons after leaving the system to atmospheric conditions. However, it must be kept in mind that GC–MS is not the most suitable technique for analyzing heavy oil samples as the amount of
Fig. 6. Phase envelope for propane with experimental pressure and temperature indicated by the red points. The numbers indicate the recovery for the corresponding experiment and the numbers in parentheses indicate the experiment number, showing that recovery is higher if the experimental conditions are very slightly on the right of the phase envelope, whereas, when the temperature is very high, the recovery falls.
5.4. Viscosity The viscosity was found to reduce, as per expectations. It was observed that the reduction in viscosity of produced oil was more in the case of butane, which corresponds with the lower asphaltene content of the oil produced with butane as solvent. To give an average value, propane reduced the viscosity by 2 to 3 times, whereas the viscosity by butane was 6 to 7 times. This can be explained by evaluating the mass transfer involved with propane and butane. For example, in experiment-3, the temperature used was 98 °C. At this pressure the saturation pressure of butane is about 1470 kPa. From Eq. 2, it corresponds to a butane mole fraction of 0.95 at equilibrium conditions. On the other hand, in experiment 11, the temperature used was 52 °C. At
Fig. 8. Phase envelope for propane with experimental pressure and temperature indicated by the red points. The numbers indicate the ratio of recovery for the corresponding experiment and sample height (in cm) also giving a value of 4–6 for the best experiments. The numbers in parentheses indicate the experiment number.
V. Pathak et al. / Journal of Petroleum Science and Engineering 78 (2011) 637–645
Butane experiments 90
Propane experiments 80 70 60 50 40 0
(T-Tsat)*P Fig. 9. Recovery change with increasing value of product of difference between experimental temperature and saturation temperature, and experimental pressure (for glass-bead experiments).
this pressure the saturation pressure of propane is 1790 kPa. From Eq. 2, it corresponds to a propane mole fraction of 0.84 at equilibrium conditions. Thus, the mole fraction of butane in the oil is higher under experimental conditions. Moreover, the molecular mass of butane is about 1.25 times higher than that of propane. This means that the mass of butane dissolved in the heavy oil is further greater than mass of propane dissolved in heavy oil. This effect is likely to contribute towards the higher viscosity reduction caused by butane. Moreover, butane experiments are done at a higher temperature as compared to propane experiments, so as to maintain butane in vapor phase. This also contributes towards the higher viscosity reduction. Results for three samples tested for viscosity of oil produced at increasing temperatures are shown in Fig. 14. The viscosity values of the produced oil are given in Table 3. For experiments with bitumen, butane reduced the viscosity from 22,000 cP to 8,000 cP whereas propane reduced the viscosity to only 16,700 cP (all values at 60 °C).
5.5. Refractive index (RI) The refractive index is a good indicator of the solvent concentration in oil and has been used in the past to predict the onset of asphaltene precipitation on addition of solvent to heavy oil (Buckley et al., 1998). In our analysis, the refractive indices were measured for all samples using a digital refractometer at a constant temperature of 25 °C. The refractive index of the sample and asphaltene content were found to be related and refractive index increased with asphaltene content (Fig. 15). This relates well with the other data and is a direct indicator of the fact that a medium with higher asphaltene content will be a denser medium, and refractive index can be used, at least qualitatively, to compare between different oil samples for asphaltene content.
Fig. 11. Setup for asphaltene measurement using ﬁlter paper and excess solvent.
5.6. Flow dynamics of the process Three possible factors affecting the recovery of heavy-oil or bitumen from sands were studied; pressure, temperature, and sample height. Their effects were clariﬁed in this paper qualitatively and quantitatively (Eqs. 3 and 4). One critical issue is indirect effects of these three parameters. Pressure, temperature and solvent type used have direct inﬂuence on asphaltene precipitation. The blockage of pores by asphaltene particles, however, is indirectly controlled by sample height. As the process is dominated by gravity drainage, the thickness of the asphaltene deposition zone has to do with the sample height (or travel time of the particles through the sample). In almost all the experiments, asphaltenic deposits were observed in the lower half of the sample after the experiments were ﬁnished. During the ﬂuid ﬂow, this deposit would have formed a lower permeability zone hindering the downward ﬂow of diluted oil. Note that this is a capillary holdup phenomenon; the height of which stays the same as the reservoir thickness increases. The results given in Fig. 5 through 8 show no speciﬁc trend even if only pressure and temperature are used (Figs. 5 and 6) or sample height is included (Figs. 7 and 8). This no-trend behavior can be attributed to other factors which indirectly impact the recovery factor, and asphaltene precipitation is the more critical one. This effect needs further clariﬁcations also supported by numerical and visual investigations. In a succeeding paper this issue has been investigated in-depth (Pathak et al., 2011). 5.7. Extraction rates Understanding the factors affecting the extraction rate in a hot solvent injection is critical for upscaling such a technique to ﬁeld scale. Several researchers studied this topic in detail and proposed correlations for
80.0 70.0 60.0 50.0 40.0 30.0
Experiment No. 3
Fig. 10. Curve ﬁt for recovery as a function of sample height, experimental temperature and pressure.
Fig. 12. Results for asphaltene content in produced oil using ﬁlter paper and excess solvent for experiments given in Table 3.
V. Pathak et al. / Journal of Petroleum Science and Engineering 78 (2011) 637–645 1.560
1.555 1.550 1.545 1.540 1.535 1.530 5.000
Asphaltene weight % in oil Fig. 15. Change of refractive index with asphaltene content of produced oil.
Fig. 13. Results for asphaltene deposition (wt.% of original oil left as precipitates in the sample after the experiment) for experiments given in Table 3. (Asphaltene left in rock = asphaltene introduced in rock — asphaltene produced with produced oil).
predicting parameters like mass ﬂux of solvent (Nenniger and Dunn, 2008). Such an analysis was beyond the scope of the current work, mainly because the experiments conducted were not continuous injectionproduction type experiments. However, we believe that the parameters affecting the overall recovery (temperature, pressure, sample height and asphaltene deposition — as discussed above) will also affect the production rate. Another parameter, which is crucial in predicting production rates, is the sample permeability. In fact, it was observed in the experiments performed, in which the soaking time of a few hours was enough to drain signiﬁcant amount of original oil from glass bead samples (permeability of the order of darcies), whereas for Berea sandstone cores (permeability of the order of millidarcies), much less oil was recovered even after a soaking time of up to 2 days. Moreover, lower permeability implies smaller pore throats and thereby higher capillary pressure, which can increase the possibility of blockage if there is some asphaltene deposition on the rock grain surface. To some extent, this issue was addressed in a succeeding paper (Pathak et al., 2011).
6. Conclusions 1. Hot solvent experiments were conducted using propane and butane and were analyzed for the recovery and properties of the produced oil. The asphaltene content of the produced oil was measured using the excess solvent (heptane)-ﬁlter paper method. Asphaltene precipitated was also inferred from this result. Butane was found to “upgrade” the oil more than propane. The asphaltene content of the oil produced with propane as a solvent was found to
Viscosity of produced oil (cp)
be higher than the asphaltene content of the oil produced with butane as a solvent. Viscosity was also higher for the oil produced with propane as a solvent as compared to the oil produced with butane as a solvent. One reason for this may be a higher amount of mass transfer of butane into the oil as compared to propane into the oil. Another reason may be that butane experiments were conducted at a higher temperature than propane experiments, which assisted in reduction of oil viscosity too. Oil recovery was observed to decrease with increase in temperature and pressure. The recovery was observed to decrease with the increase in product of temperature difference between experiment temperature and saturation temperature, and experimental pressure. The recovery was more sensitive to temperature and the most signiﬁcant change in recovery was obtained with increasing temperature. The recovery reduces if temperature is much higher than the saturation temperature at experimental pressure. This is in accordance with the Raoult's law for mixtures. A non-linear correlation between the recovery and these three quantities has been reported based on the best ﬁt of the experimental data using various functions. This correlation also shows that recovery is more sensitive to temperature and is adversely affected by increase in temperature and pressure. Increase in sample height favored higher recovery. The produced oil does not seem to have any light components as shown by the GC–MS analysis. This means that the original oil has been upgraded but still does not have any light components. However, it should be kept in mind that GC–MS is not the most suitable technique to analyze oil samples as the amount of the sample introduced into the instrument is very small and that makes it difﬁcult to detect traces of light hydrocarbons. A headspace-GC analysis is more recommended in this type of analysis as this technique involves an instrument which allows a much larger sample size to be introduced and tested.
Propane Exp 9
Butane Exp 15
6000 5000 4000 3000 2000 1000 0 20
Temperature (deg C) Fig. 14. Change in viscosity with temperature for the tested heavy oil and produced oils with propane and butane.
This research was funded by an NSERC CRD Grant (No: 385289–09) and Laricina Energy Project (No: RES0002567). The funds for the equipment used were obtained from the Canadian Foundation for Innovation (CFI) (Project # 7566) and the University of Alberta. We gratefully acknowledge these supports. We thank Dr. Zhenghe Xu and Mei Zhang for the help during viscosity measurements and Wayne Moffat for his suggestions on gas chromatography. We are also thankful to Laricina Energy for providing bitumen samples and permission to use data needed in this research and to Stephen Gamble for his valuable efforts in setting up the experimental apparatus. This paper is the revised and improved version of SPE 137440 presented at the 2010 SPE Canadian Unconventional Resources and Int. Petr. Conf., Calgary, AB, Canada, 19–21 Oct.
V. Pathak et al. / Journal of Petroleum Science and Engineering 78 (2011) 637–645
Appendix A Sample calculations for Experiments 10 and 11: Propane at 1500 psia pressure and different temperatures. Experiment 10 Ppropane at 67 ∘C ¼ 2438:09 kPa Vapor mole fraction of solvent,ypropane = 1.00 (considering oil/ bitumen to be completely non-volatile). Liquid phase mole fraction of solvent, xpropane ¼ ypropane
Ptotal 1500 ¼ 0:615 ¼1 2438:09 Ppropane
Experiment 11 Ppropane at 52 ∘C ¼ 1789:67 kPa Vapor mole fraction of solvent,ypropane = 1.00 (considering oil/bitumen to be completely non-volatile). Liquid phase mole fraction of solvent, xpropane ¼ ypropane
Ptotal 1500 ¼ 0:838 ¼1 1789:67 Ppropane
Thus, because of a higher solvent concentration in experiment 11, a higher recovery can be expected. References Allen, A.; Gillespie, R.; Burnett, D. 1984., Superheated Solvent Method for Recovering Viscous Petroleum. US Patent No. 4,450,913. Buckley, J.S., Hirasaki, G.J., Liu, Y., Von Drasek, S., Wang, J.-X., Gill, B.S., 1998. Asphaltene precipitation and solvent properties of crude oils. Journal of Petroleum Science and Technology 16 (3&4), 251–285. Butler, R.M., Mokrys, I.J., 1991. A new process (Vapex) for recovering heavy oils using hot water and hydrocarbon vapour. Journal of Canadian Petroleum Technology 30 (1), 97–106. Das, S.K., Butler, R.M., 1996. Countercurrent Extraction of Heavy Oil and Bitumen. Paper SPE 37094 presented at the SPE International Conference on Horizontal Well Technology, Calgary, Canada, 18-20 November. doi:10.2118/37094-MS.
Edmunds, N., Barrett, K., Solanki, S., Cimolai, M., Wong, A., 2009a. Prospects for commercial bitumen recovery from the Grosmont Carbonate, Alberta. Journal of Canadian Petroleum Technology 48 (9), 26–32. Edmunds, N., Moini, B., Peterson, J., 2009b. Advanced solvent-additive processes via genetic optimization. Paper PETSOC 2009–115 presented at Canadian International Petroleum Conference (CIPC) 2009, Calgary, Alberta, Canada, 16–18 June. Frauenfeld, T., Jossy, C., Wang, X., 2007. Experimental studies of thermal solvent oil recovery process for live heavy oil. Journal of Canadian Petroleum Technology 46 (1), 40–46. Friedrich, K. 2005. Effects of a Non-Condensable gas on the Vapex Process. M.Sc dissertation, MS thesis, University of Waterloo, Waterloo, Ontario, Canada. Goite, J.G., Mamora, D.D., Ferguson, M.A., 2001. Experimental study of Morichal heavy oil recovery using combined steam and propane injection. Paper SPE 69566 presented at the SPE Latin American and Caribbean Petroleum Engineering Conference, Buenos Aires, Argentina, 25–28 March. doi:10.2118/69566-MS. Haghighat, P., Maini, B.B., 2008. Role of asphaltene precipitation in Vapex process. Paper PETSOC 2008–087 presented at the Canadian International Petroleum Conference/SPE Gas Technology Symposium 2008 Joint Conference, Calgary, Alberta, Canada, June 17–19. Ivory, J., Zheng, R., Nasr, T., Deng, X., Beaulieu, G., Heck, G., 2008. Investigation of low pressure ES-SAGD. Paper SPE 117759 presented at the SPE International Thermal Operations and Heavy Oil Symposium, Calgary, Alberta, Canada, 20–23 October. Johnson, F.S., Jones, R.A., Miller, J.S., 1975. Laboratory investigation using solvent to recover heavy oil from a fractured reservoir. Paper SPE 5336 presented at the Rocky Mountain Regional Meeting of SPE of AIME, Denver, Colorado, April 7–9. Nasr, T.N., Beaulieu, G., Golbeck, H., Heck, G., 2003. Novel expanding solvent-SAGD process “ES-SAGD”. Journal of Canadian Petroleum Technology 42 (1), 13–16 (technical note). Nenniger, J.E., Dunn, S.G., 2008. How fast is solvent based gravity drainage? Paper PETSOC 2008–139 presented at the Canadian International Petroleum Conference/SPE Gas Technology Symposium 2008 Joint Conference, Calgary, Alberta, Canada, June 17–19. Palmgren, C., Edmunds, N., 1995. High temperature naptha to replace steam in the SAGD process. Paper SPE 30294 presented at the International Heavy Oil Symposium, Calgary, Alberta, Canada, 19–21 June. doi:10.2118/30294. Pathak, V., Babadagli, T., Edmunds, N.R., 2011. Mechanics of heavy oil and bitumen recovery by hot solvent injection. Paper SPE 144546, presented at the SPE Western North American Regional Meeting, Anchorage, Alaska, USA, 7–11 May. Redford, D., McKay, A., 1980. Hydrocarbon-steam processes for recovery of bitumen from oil sands. Paper SPE 8823 presented at the First Joint SPE/DOE Symposium on Enhanced Oil Recovery, Tulsa, Oklahoma, 20–23 April. Rezaei, N., Chatzis, I., 2007. Incorporation of heat in the VAPEX process: Warm VAPEX. Paper PETSOC 2007-133-EA presented at the Canadian International Petroleum Conference/Annual Technical Meeting, Calgary, Alberta, Canada, June 12–14. Shu, W.R., Hartman, K.J., 1988. Effect of solvent on steam recovery of heavy oil. SPE Res Eng. 1988, 457–564 (May). Singhal, A.K., Das, S.K., Leggitt, S.M., Kasraie, M., Ito, Y., 1996. Screening of reservoirs for exploitation by application of steam assisted gravity drainage/Vapex processes. Paper SPE 37144 presented at the International Conference on Horizontal Well Technology, Calgary, Alberta, Canada, 18–20 November. doi:10.2118/37144. Sirota, E.B., 2005. Physical structure of asphaltenes. Energy & Fuels 19, 1290–1296. Upreti, S.R., Lohi, A., Kapadia, R., El-Haj, R., 2007. Vapor extraction of heavy oil and bitumen: a review. Energy and Fuels 21, 1562–1574.